Conference televisions, as advanced communication means, have been widely used in the fields such as government meetings, distance learning, remote diagnostics and trans-regional remote conferences etc.
In the related art, a method for implementing multi-picture in a conference television is at a Multipoint Control Unit (MCU) side, wherein, a Digital Signal Processor (DSP) at the MCU side firstly decodes code streams transmitted by multiple terminals, then performs multi-picture synthesizing on the decoded data according to multi-picture layout information issued by a master control Multipoint Processor (MP) in the MCU side, and finally encodes and then outputs the synthesized image.
In the related art, if it is to implement amplifying one of multiple pictures temporarily for view, it is very inconvenient to retrieve the original image immediately, which needs to switch a multi-picture layout, i.e., changing an existing multi-picture layout A into a multi-picture layout B (in the layout B, the picture needed to be viewed is relatively large), and retrieve the multi-picture layout A after the viewing is complete.
In the processing at the MCU side, when changing the multi-picture layout A into the multi-picture layout B, the master control MP needs to issue the layout B to the DSP and the DSP to compute multi-picture related information of the layout B (initial coordinates, width, height etc. of each picture in the multiple pictures) and to perform corresponding multi-picture synthesizing, and then encode and output; and the master control MP needs to re-issue the layout A to the DSP and the DSP re-computes multi-picture information of the layout A and performs corresponding multi-picture synthesizing.
Disadvantages of the above method are in that: first, the processing in the MCU is complex, and there are multiple repeated computations; and second, the image needed to be amplified cannot be displayed flexibly, and can only be selected according to the current layout.